There was a lively debate regarding extinction. Of course, we’re talking about behavioral extinction, not extinction of an entire species. During the long and heated debate, it was claimed that at least one prominent trainer is teaching that there is no such thing as extinction. Seriously?
The threat of extinction is real, yet the uncertainty is reasonable. There is a punishment method that seems to be pretty comparable until you dig a little further. It is critical to distinguish between these processes, therefore for the purpose of clarity, I will return to the foundations and begin from there. Please follow me.
To comprehend extinction, it is necessary to first grasp what reinforcement is, therefore I will begin there.
The fundamentals of reinforcement
Reinforcement is the process through which a behavior is followed by an event that enhances or maintains the future rate of that behavior. The process is only reinforced if the rate of the behavior rises or remains constant as a result of the event that preceded it. If it works, the occurrence is known as a reinforcer. The event is only a reinforcer if it occurs after the behavior and raises or maintains the pace of the behavior. Food, beverage, an opportunity to leave an uncomfortable circumstance, some social engagement, or a number of other things might be the reinforcing event.
The reinforcing event must also occur as a result of that conduct. By contingent, I mean that the reinforcer must occur if and only if the desired behavior occurs. The reinforcer and the behavior are interconnected.
Small quantities of food are the most often used reinforcer in clicker training. When an animal performs a behavior, the trainer clicks the clicker and creates a reward, causing the animal to repeat the activity in the future.
What is the relationship between extinction and reinforcement?
Although extinction refers to a decrease in the frequency of a behavior, it can only occur when reinforcement has previously been used to keep the habit in place. Extinction is sometimes defined as just ignoring a behavior until it disappears. That is not really correct, but it may seem such.
Extinction is the process by which the rate of a behavior declines because a reinforcer associated with that behavior is no longer given. In other words, a behavior is rewarded for a while and its rate grows or remains steady, but subsequently the reinforcer is no longer supplied when the behavior occurs, and the rate of the activity decreases. The behavior drops during the extinction process because the reinforcer is no longer supplied when the behavior occurs. That dependent link between the reinforcer and the behavior no longer exists. They’ve split up. Jennifer is no longer present wherever Brad is. Everything has changed. As a consequence, the behavior reduces.
This is where the misunderstanding arises.
That is what extinction is. But there is another mechanism that sounds similar to extinction. At first glance, negative punishment sounds so similar to extinction that some people believe they are the same thing. But they disregard it too quickly.
Negative punishment is a process in which an event is taken from the environment as a result of a behavior, and the rate of that conduct lowers in the future. That sounds a lot like extinction, which is the process through which a reinforcer is no longer provided and the rate of behavior drops. Something is missing in both cases, and the rate of a behavior drops in both cases. However, there are several key differences.
In both circumstances, an operant behavior is taking place. That signifies the behavior is taking place because the reinforcer is linked to it. (Operant is a word used by B.F. Skinner to describe how a behavior interacts with its surroundings. In these circumstances, the conduct affects the environment in order to provide some reinforcer.) As a result, when the shift occurs, these habits are rewarded.
Extinction
The behavior involves interacting with the environment in order to provide a reinforcer. However, when the behavior occurs, the reinforcer is no longer supplied, and the behavior ceases. Because the reinforcer was not easily accessible in the environment until the activity occurred, the behavior acted on the environment to create it. The behavior then reduced since it no longer had the impact of creating a reinforcer. It is critical to notice that the item that is terminated in extinction is the precise thing that kept the behavior going in the first place—the reinforcer.
When Fido initially came to live with Janie, she gave him a reward every time he stood on his rear legs, so Fido did this a lot. Janie ran out of snacks after a few weeks and didn’t bother getting any more. Fido continued to stand on his rear legs around her for a time, but it became less and less frequent. When the goodies for his standing behavior were no longer supplied, his standing behavior died.
The goodies made him stand on his back legs, and when the sweets stopped coming, he eventually stopped standing on his back legs.
Punishment in the Negative
Negative punishment operates on the environment to provide some reinforcer to begin with, similar to extinction. When the activity occurs, however, something else that was easily accessible in the environment gets removed from the environment. If the action is completed, the reinforcer remains accessible, but the loss of this other item causes the behavior to occur less often in the future. It is not the discontinuance of the reinforcer, as in extinction, that triggers the change, but rather the loss of something else.
Jimmy shrieked every time he pulled his sister Sally’s hair, so he pulled her hair a lot. Their mother had had enough of this, so when Jimmy grabbed Sally’s hair one day, she confiscated his Game Cube system and put it in the trunk of the vehicle. Jimmy soon got his Game Cube back, and he started pulling Sally’s hair less often (at least in front of Mom) so he wouldn’t lose it again.
In contrast to extinction, the reinforcer for yanking Sally’s hair was still present… she would have shrieked when he pulled her hair. But Jimmy lost something important when he ripped her hair. Even though the Game Cube had nothing to do with his pulling Sally’s hair in the first place, its absence caused him to pull Sally’s hair less often in the future.
Why it is significant
In nature, behavioral processes such as negative and positive reward, negative and positive punishment, and extinction occur. They are natural processes, similar to those found in physics, chemistry, and other disciplines. They do not occur simply because trainers “do” them to learners. If a bear visits a certain berry patch for days and days and is rewarded with berries to eat, but the berries eventually run out and he gradually stops hunting for berries there, that is extinction. It would also be extinction if Bruce stopped stroking his dog, Milo, every time he jumped on him, and Milo progressively stopped leaping on him. It is a natural process that trainers may learn to use to their advantage, much as pharmacists use chemical rules to create medications.
The same may be said about negative punishment. If a newborn fox rushes off gathering bugs but finds that his mother is not to be seen when he turns around, he may go bug catching less often in the future, at least until he is a little older. His mother, who had always been around, has vanished because he pursued a bug. The reinforcer for bug capturing is still accessible… either the joy of the chase or the eating of the insect, but his bug pursuing is adversely penalized by Mom’s absence. Similarly, the human Mom may take away Jimmy’s Game Cube to urge him to stop tugging Sally’s hair, which is negative punishment if it succeeds.
The crucial thing for trainers to remember is that we need to know what process is in place before we can successfully modify it in order to influence behavior.
By removing extinction from our training vocabulary, we lose a viable reason for failing behavior and a crucial component of an exceedingly powerful behavior-building tool, differential reinforcement.
Extinction is required for shaping and differential reinforcing.
Differential reinforcement is a component of shaping, which is important to clicker training. Differential reinforcement involves reinforcing the new approach to the goal behavior while allowing the prior approximation fade away. The differential reinforcement of successive approximations toward a desired behavior is referred to as shaping. And what exactly is extinction? It is the process through which the rate of a behavior decreases when a reinforcer is no longer given with it. It would be quite difficult to locate negative punishers that would properly assist you shape toward new habits. When it comes to extinction, you already know what to do. You know that when you employ a given treat or other reinforcer, your learner’s behavior rate rises, and you also know that when you cease giving that reinforcer, the rate of that behavior decreases. So you alter the criteria for obtaining that reinforcer with care. The reinforcer contains the magic… and in the trainer’s molding dancing ability.
Differential reinforcement is also an effective method for reducing undesirable behaviour. A frequent method is to encourage the heck out of a preferred behavior while extinguishing the undesirable behavior. It is especially helpful if the same reinforcer that was used to sustain the problem behavior is used to reinforce the desired behavior. So, if Fido leaps on family members when they get home in order to gain attention and petting, the family may cease giving attention for leaping and only provide attention when Fido sits at their feet. Fido will sit at their feet considerably more frequently than he will jump on them as the training process goes if they are consistent.
One of the most appealing aspects of differential reinforcement is the combination of reward for one behavior and extinction for another… is that you can often avoid or eliminate the harmful extinction spurts that might occur when using extinction alone. You may frequently prevent or significantly limit extinction bursts by giving a desired option for the learner to acquire the reinforcer he has previously shown he will work for. However, you still receive enough of a burst during shaping to obtain some unpredictable behavior, giving you additional estimates to work with.
The final conclusion is that extinction is a true process that you should comprehend… and believe in!
